comparison preprocessing.xml @ 4:141a9288be9c draft

planemo upload for repository https://github.com/galaxyproteomics/tools-galaxyp/tree/master/tools/cardinal commit ecdc3a64aa245d80dbc5487b2bf10a85a43adc6d

3:f172efe92629

<tool id="cardinal_preprocessing" name="MSI preprocessing" version="@VERSION@.2">
    <description>
        mass spectrometry imaging preprocessing
    </description>
    <macros>
        <import>macros.xml</import>

library(lattice)
library(ggplot2)

@READING_MSIDATA@

## remove duplicated coordinates, otherwise peak picking and log2 transformation will fail
print(paste0(sum(duplicated(coord(msidata)[,1:2])), " duplicated coordinates were removed"))
msidata <- msidata[,!duplicated(coord(msidata)[,1:2])]


if (ncol(msidata)>0 & nrow(msidata) >0){                                         

            #elif str( $method.methods_conditional.methods_for_smoothing.smoothing_method) == 'sgolay':
                print('sgolay smoothing')

                msidata = smoothSignal(msidata, method="$method.methods_conditional.methods_for_smoothing.smoothing_method", window=$method.methods_conditional.window_smoothing, order = $method.methods_conditional.methods_for_smoothing.order_of_filters)
            #elif str($method.methods_conditional.methods_for_smoothing.smoothing_method) == 'ma':
                print('moving average smoothing')

                msidata = smoothSignal(msidata, method="$method.methods_conditional.methods_for_smoothing.smoothing_method", window=$method.methods_conditional.window_smoothing, coef = $method.methods_conditional.methods_for_smoothing.coefficients_ma_filter)

                msidata = reduceDimension(msidata, method="bin", width=$method.methods_conditional.methods_for_reduction.bin_width, units="$method.methods_conditional.methods_for_reduction.bin_units", fun=$method.methods_conditional.methods_for_reduction.bin_fun)

                ## optional: replace NA with 0
                #if $method.methods_conditional.methods_for_reduction.replace_NA_bin:
                    ## binning seems to create normal R matrix but to be sure:
                    iData(msidata) <- iData(msidata)[]
                    ## count and replace NAs
                    print(paste0("Number of NA that were set to zero after binning:",sum(is.na(spectra(msidata)[]))))
                    spectra(msidata)[is.na(spectra(msidata))] = 0 
                #end if

            #elif str( $method.methods_conditional.methods_for_reduction.reduction_method) == 'resample':
                print('resample reduction')

        #elif str( $method.methods_conditional.preprocessing_method) == 'Transformation':
            print('Transformation')

            ## convert data into R matrix what brings it automatically into memory and can take some take but next steps need R matrix
            iData(msidata) <- iData(msidata)[]

            #if str( $method.methods_conditional.transf_conditional.trans_type) == 'log2':
                print('log2 transformation')

                ## replace 0 with NA to prevent Inf
                spectra_df = spectra(msidata)[]
                spectra_df[spectra_df ==0] = NA
                print(paste0("Number of 0 which were converted into NA:",sum(is.na(spectra_df))))
                spectra(msidata) = spectra_df
                ## log transformation
                spectra(msidata) = log2(spectra(msidata))
                ## optional: replace NA with 0
                #if $method.methods_conditional.transf_conditional.replace_NA_trans:
                    spectra(msidata)[][is.na(spectra(msidata)[])] = 0
                #end if

            #elif str( $method.methods_conditional.transf_conditional.trans_type) == 'sqrt':
                print('squareroot transformation')

                spectra(msidata)[] = sqrt(spectra(msidata)[])

           #end if

            ############################### QC ###########################

                            <option value="sgolay">Savitsky-Golay</option>
                            <option value="ma">moving average</option>
                        </param>
                        <when value="gaussian">
                            <param name="sd_gaussian" type="float" value="2"
                                   label="The standard deviation for the Gaussian kernel (window/sd)"/>
                        </when>
                        <when value="sgolay">
                            <param name="order_of_filters" type="integer" value="3"
                                   label="The order of the smoothing filter, must be smaller than window size"/>
                        </when>
                        <when value="ma">
                            <param name="coefficients_ma_filter" type="float" value="1"
                                   label="The coefficients for the moving average filter"/>
                        </when>

                            <param name="bin_fun" type="select" display="radio"
                                   label="Calculate sum or mean intensity for ions of the same bin">
                                    <option value="mean" selected="True">mean</option>
                                    <option value="sum">sum</option>
                            </param>
                            <param name="replace_NA_bin" type="boolean" label="Replace NA with 0" truevalue="TRUE" falsevalue="FALSE" help="Binning can introduce NAs, should they be replaced with 0"/>
                        </when>
                        <when value="resample">
                            <param name="resample_step" type="float" value="1"
                                   label="The step size in m/z" help="Step size must be greater than range of m/z values divided by number of m/z features"/>
                        </when>

                        <param name="trans_type" type="select" label="Intensity transformations" help="logarithm base 2 (log2) or squareroot (sqrt)">
                            <option value="log2" selected="True">log2</option>
                            <option value="sqrt">sqrt</option>
                        </param>
                            <when value="log2">
                                <param name="replace_NA_trans" type="boolean" label="Replace NA with 0" truevalue="TRUE" falsevalue="FALSE" help="0 values are set to NA before log2 transformation, after transformation they can be set back to 0"/>
                            </when>
                            <when value="sqrt"/>
                    </conditional>
                </when>
            </conditional>
        </repeat>
        <param name="imzml_output" type="select" display = "radio" optional = "False"
               label="Output format" help= "Choose the output format">
                <option value="imzml_format" selected="True">imzML</option>
                <option value="rdata_format">RData</option>
        </param>
    </inputs>
    <outputs>
        <data format="imzml" name="outfile_imzml" label="${tool.name} on ${on_string}: imzML">
            <filter>imzml_output=='imzml_format'</filter>

- Peak filtering works only on centroided data (peak picking and alignment or Data reduction peaks

**Output**

- MSI data as imzML file or .RData (can be read with the Cardinal package in R)
- pdf with key values and average mass spectra after each processing step

        ]]>
    </help>
    <expand macro="citations"/>
</tool>

4:141a9288be9c

<tool id="cardinal_preprocessing" name="MSI preprocessing" version="@VERSION@.3">
    <description>
        mass spectrometry imaging preprocessing
    </description>
    <macros>
        <import>macros.xml</import>

library(lattice)
library(ggplot2)

@READING_MSIDATA@

@READING_MSIDATA_INRAM@  ###change out

## remove duplicated coordinates, otherwise peak picking and log2 transformation will fail
msidata <- msidata[,!duplicated(coord(msidata)[,1:2])]


if (ncol(msidata)>0 & nrow(msidata) >0){                                         

            #elif str( $method.methods_conditional.methods_for_smoothing.smoothing_method) == 'sgolay':
                print('sgolay smoothing')

                msidata = smoothSignal(msidata, method="$method.methods_conditional.methods_for_smoothing.smoothing_method", window=$method.methods_conditional.window_smoothing, order = $method.methods_conditional.methods_for_smoothing.order_of_filters)

                ## if selected replace negative intensities with zero
                #if $method.methods_conditional.methods_for_smoothing.replace_negatives:
                    spectra(msidata)[spectra(msidata)<0] = 0
                #end if

            #elif str($method.methods_conditional.methods_for_smoothing.smoothing_method) == 'ma':
                print('moving average smoothing')

                msidata = smoothSignal(msidata, method="$method.methods_conditional.methods_for_smoothing.smoothing_method", window=$method.methods_conditional.window_smoothing, coef = $method.methods_conditional.methods_for_smoothing.coefficients_ma_filter)

                msidata = reduceDimension(msidata, method="bin", width=$method.methods_conditional.methods_for_reduction.bin_width, units="$method.methods_conditional.methods_for_reduction.bin_units", fun=$method.methods_conditional.methods_for_reduction.bin_fun)

                ## optional: replace NA with 0
                #if $method.methods_conditional.methods_for_reduction.replace_NA_bin:
                    ## count and replace NAs
                    print(paste0("Number of NA that were set to zero after binning:",sum(is.na(spectra(msidata)))))
                    spectra(msidata)[is.na(spectra(msidata))] = 0 
                #end if

            #elif str( $method.methods_conditional.methods_for_reduction.reduction_method) == 'resample':
                print('resample reduction')

        #elif str( $method.methods_conditional.preprocessing_method) == 'Transformation':
            print('Transformation')

            ## convert data into R matrix what brings it automatically into memory and can take some take but next steps need R matrix
            ##iData(msidata) <- iData(msidata)[]

            #if str( $method.methods_conditional.transf_conditional.trans_type) == 'log2':
                print('log2 transformation')

                ## replace 0 with NA to prevent Inf
                spectra_df = spectra(msidata)[] ## convert into R matrix
                spectra_df[spectra_df ==0] = NA
                print(paste0("Number of 0 which were converted into NA:",sum(is.na(spectra_df))))
                spectra(msidata) = spectra_df
                ## log transformation
                spectra(msidata) = log2(spectra(msidata))
                ## optional: replace NA with 0
                #if $method.methods_conditional.transf_conditional.replace_NA_trans:
                    spectra(msidata)[is.na(spectra(msidata))] = 0
                #end if

            #elif str( $method.methods_conditional.transf_conditional.trans_type) == 'sqrt':
                print('squareroot transformation')

                spectra(msidata) = sqrt(spectra(msidata)[])

           #end if

            ############################### QC ###########################

                            <option value="sgolay">Savitsky-Golay</option>
                            <option value="ma">moving average</option>
                        </param>
                        <when value="gaussian">
                            <param name="sd_gaussian" type="float" value="2"
                                   label="The standard deviation for the Gaussian kernel. Default = window/4"/>
                        </when>
                        <when value="sgolay">
                            <param name="order_of_filters" type="integer" value="3"
                                   label="The order of the smoothing filter, must be smaller than window size"/>
                            <param name="replace_negatives" type="boolean" label="Replace negative intensities with 0" truevalue="TRUE" falsevalue="FALSE" checked="True" help="Savitzky golay smoothing can introduce negative intensity values it is recommended to replace them with zero"/>
                        </when>
                        <when value="ma">
                            <param name="coefficients_ma_filter" type="float" value="1"
                                   label="The coefficients for the moving average filter"/>
                        </when>

                            <param name="bin_fun" type="select" display="radio"
                                   label="Calculate sum or mean intensity for ions of the same bin">
                                    <option value="mean" selected="True">mean</option>
                                    <option value="sum">sum</option>
                            </param>
                            <param name="replace_NA_bin" type="boolean" label="Replace NA with 0" truevalue="TRUE" falsevalue="FALSE" checked="True" help="Binning can introduce NAs, should they be replaced with 0"/>
                        </when>
                        <when value="resample">
                            <param name="resample_step" type="float" value="1"
                                   label="The step size in m/z" help="Step size must be greater than range of m/z values divided by number of m/z features"/>
                        </when>

                        <param name="trans_type" type="select" label="Intensity transformations" help="logarithm base 2 (log2) or squareroot (sqrt)">
                            <option value="log2" selected="True">log2</option>
                            <option value="sqrt">sqrt</option>
                        </param>
                            <when value="log2">
                                <param name="replace_NA_trans" type="boolean" label="Replace NA with 0" truevalue="TRUE" falsevalue="FALSE" checked="True" help="0 values are set to NA before log2 transformation, after transformation they can be set back to 0"/>
                            </when>
                            <when value="sqrt"/>
                    </conditional>
                </when>
            </conditional>
        </repeat>
        <param name="imzml_output" type="select" display = "radio" optional = "False"
               label="Output format" help= "Choose the output format">
                <option value="imzml_format" >imzML</option>
                <option value="rdata_format" selected="True" >RData</option>
        </param>
    </inputs>
    <outputs>
        <data format="imzml" name="outfile_imzml" label="${tool.name} on ${on_string}: imzML">
            <filter>imzml_output=='imzml_format'</filter>

- Peak filtering works only on centroided data (peak picking and alignment or Data reduction peaks

**Output**

- MSI data as imzML file or .RData (can be read with the Cardinal package in R)
- pdf with key values and four random mass spectra after each processing step

        ]]>
    </help>
    <expand macro="citations"/>
</tool>